536 REPORT— 1892. 



Familiarity -with the subaqueous workings had, however, only served to 

 confirm the author's opinion of the feasibility of the undertaking, which 

 at an earlier stage he had offered to carry out without the intervention of 

 a contractor, and this course was now decided upon. 



To have to conduct the work with a shield insufficiently strong and 

 in such a condition was not a satisfactory beginning. The author had no 

 doubt whatever that with a properly designed shield it could have been 

 rapidly performed. He was not certain that it could be done with the 

 existing shield, but all things considered, and rather than attempt to 

 replace it in the disturbed ground in which it then lay, it was thought 

 better to repair and strengthen the existing shield, and at least to make 

 an effort to proceed with it. The repairing and strengthening of the 

 cutting edge was successfully performed, and 42^ months after the com- 

 mencement of the first contractor's work the tunnel was recommenced. 

 The most effective method was found to be to remove the lower part of 

 the diaphragm, as shown in the model exhibited, to regulate the air 

 pressure as nicely as possible at such a point below the hydrostatic 

 pressure that, while — so long as the shield was kept moving — it was 

 generally insufficient to blow a funnel from the upper edge of the 

 shield to the river above, it was just sufficient to prevent an inflow 

 at the lower edge of the shield sufficiently rapid to carry the ground 

 with it. 



At intervals round the cutting edge steel teeth were added before the 

 work was recommenced, and the author is disposed to think that they 

 were of great use in penetrating shingle. Several feet behind the main 

 diaphragm was placed a partition projecting upwards to a level above 

 that of the underside of the remaining portion of the main diaphragm. 

 This proved to be exceedingly useful when a blow of sand and water oc- 

 curred. It did not materially hinder access to or from the front of the 

 shield, while there can be no doubt that when a blow took place it choked 

 itself more readily than if such an obstruction had not existed. The 

 author was indebted to Sir Benjamin Baker for suggesting this device, 

 and that of the teeth. 



During the latter part of the progress the longitudinal crack already 

 mentioned near the bottom of the shield became worse, another crack 

 worked from it circumferentially behind the diaphragm, and the plates of 

 that portion turned down like a ploughshare from their proper position. 

 This caused a delay of twelve days, included in the 4h months. The 

 ground at the time was as bad as it could be, but the piece, 6^ feet long 

 and 9 feet wide, was successfully cut away and replaced in situ. The por- 

 tion cut av.ray could not be removed, but was left behind by the advancing 

 shield when the work was recommenced. 



In judging of this work it is not to be supposed that it was under- 

 taken in a foolhardy manner. Such an operation was never contemplated 

 when the parliamentary plans were deposited, but, on the contrary, the 

 aqueduct jnpes were shown by those plans to be laid in the bed of the 

 river, and the softest bed was chosen for the purpose — the very worst 

 place, in fact, for the construction of a tunnel. For reasons which it is 

 not necessary to recount, the question of the depth at which the aqueduct 

 should be laid was referred to the Board of Trade. The Board would 

 probably have been more merciful if it had foreseen, as the advisers of 

 the Corporation of Liverpool foresaw, the pending difficulties ; but, as a 

 matter of fact, the decision required the aqueduct to be laid at such a 



